Potassium responses to sodium zirconium cyclosilicate in hyperkalemic hemodialysis patients: post-hoc analysis of DIALIZE

Patients with end-stage kidney disease (ESKD) have severely reduced renal potassium excretion. Excessively high serum potassium (sK⁺) is known to promote ventricular arrhythmias and cardiac death; however, even modest increases, such as > 5.5 mmol/L, are associated with increased all-cause mortality and hospitalization [1‐4].

Patients with ESKD require dialysis to remove accumulated potassium, with a goal of maintaining or restoring sK⁺ to within the normal range. Maintenance hemodialysis is typically performed three times weekly, with two 1-day intervals (i.e., the short interdialytic interval [SIDI]) and one 2-day interval (the long interdialytic interval [LIDI]) between dialysis sessions [5]. During dialysis, potassium has the potential to move freely across the dialyzer membrane, typically down a concentration gradient from a patient’s blood into the dialysate [6]. Therefore, dialysate potassium (dK⁺) concentration is a modifiable factor that affects the sK⁺ concentration achieved during hemodialysis [7, 8]. Studies to date on the impact of low dK⁺ baths (typically defined as 0–1 mmol/L) on hemodialysis outcomes have yielded contrasting results. Some studies have suggested no increased risk with low dK⁺ baths [1, 9], while others have reported an increased risk of mortality, sudden death, or sudden cardiac arrest [4, 10‐13].

The potassium gradient results from the difference between a patient’s sK⁺ and dK⁺ concentrations. A low gradient among patients with high sK⁺ may result in insufficient potassium removal during dialysis, which is associated with a greater risk of hyperkalemia and increased mortality [1, 4, 13]. Dialysis with a high gradient, resulting from a high sK⁺ and/or low dK⁺, leads to greater removal of potassium than with a lower gradient [7] and has utility in controlling hyperkalemia among patients with high sK⁺ [1]. However, a high potassium gradient at the start of hemodialysis also causes more rapid fluxes of potassium and is associated with a greater risk of adverse events (AEs), such as cardiac arrhythmia, mortality, and hospitalization [1, 13‐16]. Therefore, physicians face a challenge to balance the need to remove sufficient potassium to avoid hyperkalemia, while minimizing the risk of AEs.

Hemodialysis patients with hyperkalemia often depend on additional strategies to manage hyperkalemia, typically including dietary counseling, therapies such as potassium binders or, in more extreme situations, longer or more frequent dialysis. Sodium zirconium cyclosilicate (SZC) is a novel, highly selective potassium binder that is not absorbed or metabolized by the body [17]. SZC preferentially captures potassium in the gastrointestinal lumen, thereby reducing potassium absorption and increasing potassium fecal excretion, and reducing sK⁺ [17‐19]. The efficacy and safety of SZC for the treatment of hyperkalemia have been demonstrated in phase 2 and 3 studies of non-dialysis populations with chronic kidney disease, heart failure, and diabetes [18‐23], as well as in hemodialysis patients [24]. The phase 3b DIALIZE study demonstrated that SZC is an effective and well-tolerated treatment for hyperkalemia in patients with ESKD undergoing maintenance hemodialysis [24]. Overall, 41.2% of patients receiving SZC and 1.0% of patients receiving placebo were deemed to be treatment responders [24].

Here, we report the results of post-hoc analyses of the DIALIZE data performed to further examine the spectrum of potassium responses to SZC, in terms of sK⁺ control and potassium gradient, in hyperkalemic hemodialysis patients.

Recently, Kidney Disease: Improving Global Outcomes (KDIGO) recommended further research determining the efficacy of newer agents in patients with ESKD receiving maintenance hemodialysis [25]. We used data from the phase 3b DIALIZE study to further investigate the spectrum of potassium responses with SZC in maintenance hemodialysis patients with hyperkalemia. SZC was associated with a greater proportion of patients achieving clinically recommended or acceptable pre-dialysis sK⁺ ranges and lower potassium gradient during the evaluation period versus placebo.

Our findings extend those previously reported in the phase 3b DIALIZE study regarding control of sK⁺ with SZC [24]. We assessed control of hyperkalemia using a pre-dialysis sK⁺ range of 4.0–5.0 mmol/L and an extended range of 4.0–5.5 mmol/L. Although UK clinical practice guidelines recommend maintaining a pre-dialysis sK⁺ concentration of 4.0–6.0 mmol/L [26], previous analyses have shown that sK⁺ concentrations of ≥5.6 mmol/L are associated with increased all-cause and cardiovascular mortality [1, 4, 27], and ≥ 5.5 mmol/L is associated with hospitalization [2]. For many physicians and patients, achieving and maintaining a pre-dialysis sK⁺ concentration of 4.0–5.5 mmol/L would be considered a successful response in hemodialysis patients with hyperkalemia, while preventing patients from being exposed to the risks associated with sK⁺ > 5.5 mmol/L. In our analyses, regardless of the sK⁺ range used, SZC was associated with greater control of sK⁺ versus placebo. Using the extended range of 4.0–5.5 mmol/L, nearly half of the patients receiving SZC achieved and maintained the sK⁺ range at all 4 LIDI visits over a 4-week period. Some reduction from baseline in sK⁺ was observed with placebo, which may be attributable to some patients having temporary increases in sK⁺ or to a change in patient behavior such as an increased compliance to diet restrictions.

Reports suggest that a high potassium gradient at the start of hemodialysis is associated with risk of AEs, such as cardiac arrhythmia, mortality, and hospitalization [1, 13‐16], probably related to more rapid and larger potassium fluxes during the dialysis session. For instance, Brunelli et al. used a reference potassium gradient of 2 to < 3 mmol/L, and observed an 8, 26, and 59% higher adjusted risk of hospitalization following dialysis with gradient categories of 3 to < 4, 4 to < 5, and ≥ 5 mmol/L, respectively [16]. In addition, the authors observed an increased risk of emergency department visits following dialysis of 6, 17, and 54% with potassium gradient categories of 3 to < 4, 4 to < 5, and ≥ 5 mmol/L, respectively, versus 2 to < 3 mmol/L. Although there was no significant association between risk of mortality and higher potassium gradient, the authors observed a non-significant trend towards a greater risk of cardiovascular hospitalization with ≥5 mmol/L versus 2 to < 3 mmol/L [16]. This led the authors to suggest that the observed associations may be driven by cardiac arrhythmias and their immediate consequences [16]. As a reduction in potassium gradient is often recommended [14, 15], the reduction in potassium gradient to < 3.0 mmol/L with SZC observed in our analyses could potentially lower the risks associated with these factors, although further investigation is required to confirm this.

Practice patterns of dK⁺ concentration are known to vary globally, reflecting the lack of consensus and guidelines on ideal practice. In an analysis of the Dialysis Outcomes and Practice Patterns Study, dK⁺ of 2.0–2.5 mmol/L was most commonly used worldwide, and was prescribed to 75% of patients in the US and > 99% of patients in Japan [4]. In the US, patients are typically prescribed a dK⁺ of 2.0–4.0 mmol/L [4, 16, 28], with approximately 3% of patients receiving 1.0–1.5 mmol/L [4]. Recently, a trend has been observed for steadily increasing dK⁺ concentrations in North America, Europe, Australia, and New Zealand [4], possibly resulting from increased interest in the impact of dK⁺ on hemodialysis outcomes. Several reports have shown that dK⁺ concentrations of 0–1, < 1.5, or < 2 mmol/L are associated with a higher risk of cardiac arrest, sudden death, and all-cause mortality among hemodialysis patients than with higher dK⁺ [10‐12]. Furthermore, Karaboyas et al. showed that a dK⁺ concentration of 1.0–1.5 mmol/L is associated with a higher risk of mortality than with 2.0–2.5 mmol/L [4]. Meanwhile, no meaningful differences in clinical outcomes were observed with a dK⁺ concentration of 3.0 versus 2.0 mmol/L [4, 13]. Despite these findings, dK⁺ prescriptions of 1.0–1.5 mmol/L or lower are still used; e.g. 1.0–1.5 mmol/L is used by 62% of patients in Spain and 9% of patients in the Gulf Cooperation Council [4]. Indeed, dK⁺ concentration should be no lower than is necessary to achieve potassium control, and recent publications have recommended avoiding dK⁺ of < 2.0 mmol/L [4, 12, 28‐31].

Approaches other than modifying hemodialysis factors, such as pharmacological options and education on dietary potassium sources, may merit further attention to improve hemodialysis outcomes [4, 13, 16, 28]. Concentrations of sK⁺ are typically monitored once a month in clinical practice, but more frequent monitoring of pre-dialysis sK⁺ may allow appropriate adjustment of dK⁺ concentration [1, 12, 31]. In DIALIZE, dK⁺ modification was based on investigator’s choice and locally accepted clinical practice [24]. As demonstrated in the present analyses, the potential beneficial effect on potassium gradient is due to the reduction in pre-dialysis sK⁺ with SZC, since so few patients had an increase in dK⁺ during the study. In the long-term, this may result in the ability to relax dietary potassium restrictions allowing a diet that is richer in nutrients and fiber, as well as a potential reduction in additional acute dialysis treatments, which could improve patients’ quality of life; however, this requires further investigation.

The present analyses have several limitations. Although the analyses provide interesting findings requiring further study, they are post-hoc in nature and were not prespecified; therefore, the results are exploratory and hypothesis-generating. Patient numbers within the potassium gradient categories were small, limiting the interpretation of these findings. Target sK⁺ varies in clinical practice according to local guidelines; as such, the ranges analyzed may not be applicable worldwide. The changes in potassium gradient seen were predominantly due to reduction in pre-dialysis sK⁺ with SZC, and so the potential benefit of increasing dK⁺ baths could not be explored. Finally, clinical characteristics at baseline that may contribute to control of hyperkalemia or reduction in potassium gradient were not explored. However, the analyses presented here have several strengths. Specifically, the analyses are derived from a randomized controlled trial with a robust methodology that included blinding, multiple study centers, and repeated measures during several LIDI visits.

In conclusion, these analyses expand our knowledge of the spectrum of potassium response with SZC in maintenance hemodialysis patients with hyperkalemia. Our findings suggest that treatment with SZC improves control of hyperkalemia in maintenance hemodialysis patients with hyperkalemia. A reduction in potassium gradient towards values below the reported higher risk of > 3.0 mmol/L with SZC was observed largely without changing dK⁺. Further investigation is required to determine whether this finding could potentially modify the risks associated with a high potassium gradient.